Multiple studies have failed to produce a reliable method for rapidly identifying serious bacterial infection in febrile young infants. It is eight to nine times more likely that fever that occurs in young infants results from a benign viral illness, but the potential lethality of serious bacterial infection has resulted in a standard of care that is expensive and aggressive. The studies done to detect serious bacterial infection in febrile infants include blood cultures, urine and cerebrospinal fluid cultures, and the administration of broad-spectrum antibiotics. Hospitalization for two or three days is recommended for all infants under 28 days of age. This study evaluates the utility of polymerase chain reaction (PCR)-based methods to detect the presence of pathogenic viruses of various virus families in febrile infants. Coupled with a thorough clinical evaluation, the investigator hypothesizes that the length of hospital stay will be shortened if viral detection methods prove positive. This study was presented at the site visit evaluating the competing renewal of this grant last fall. Reviewers pointed out difficulties with primary and secondary endpoints of the study, and recommended no further support from the GCRC grant for this study. A carefully conceived rebuttal was prepared and forwarded to the NIH, but this study ceased in late September of 1999. A drastically revised study answering all issues raised by reviewers was submitted to the Advisory Committee of the Utah GCRC at its meeting in January 2000. The revised application was carefully reviewed and unanimously approved. The study now has somewhat different endpoints, utilizes quite different methods, has a slightly different primary emphasis, and a new title. Dr. Byington's new study will be reviewed in the Annual Report next year.